J. Org. Chem., Vol. 39, No. 13, 1974 1963
Notes Molecular Rearrangements i n the Course of Ritter Reactions Susan Edwards and Fritz-Hans Marquardt*l
Research Department, Pharmaceutical Division, CIBA-GEIGY Corporation, Summit, New Jersey 07901 Receiued November 27, 1973
It is well known that the Ritter reaction occurs uia ionic intermediates and that molecular rearrangements take place if the initial carbonium ion intermediate can isomerize to a more stable 0118.2 Thus Boltze and Muhlenbein3 reported that a-isopropylbenzyl alcohol (1) reacts with hydrogen cyanide or with nitriles under acidic conditions to yield amides of a,a-dimethylphenethylamine(4). On the other hand, this seems to contradict the report by Christol, e t u L . , ~ that Ritter reactions of 1 form amides of the unrearranged a-isopropylbenzylamine (2). CO-R OH
I
- C6H5CH-sH NH
CH;
I I CbHjCH-CH I 1
CH3
I
CHJ
\
CHJ
2
CH3
I C6HjCH=C I CH,
I
CH, --+
I
C,H,CH,CNHCOR
I
CHs 4
3 R=CH3 b, R = C,H, a,
A check of the reaction conditions which were used by these two groups showed that a reasonable cause for this discrepancy could have been the different order of addition of reagents used, since this in turn could have easily changed the type of mechanistic factor which controlled the reactions. Thus, Boltze and Muhlenbein first mixed the carbinol and the acids, and only later added the nitrile, whereas Christol and coworkers first mixed the carbinol with the nitrile, and added the sulfuric acid last. In the first case, the initially formed carbonium ion may therefore have had ample time to rearrange into a more stable structure before the nitrile was added ( e . g . , thermodynamic control of reaction), while in the second case the nitrile could trap the carbonium ions as they were being formed ( e . g . , kinetic control). This would also explain Christol's additional observation that, while the Ritter reaction of 1 yielded 2, the same reaction with the styrene derivative 3 only yielded 4. To check our rationalization, we only changed the order of addition of reagents when 1 was allowed to react with nitriles in a mixture of sulfuric and acetic acid under otherwise identical conditions. As was expected, 2 was obtained when sulfuric acid was the last reagent which was added, while 4 was obtained when the nitrile was added last.5 Experimental Section Melting points were determined in capillaries and are uncorrected; nmr spectra were recorded on a Varian Associates A-60 spectrometer with TMS as internal standard. a-Isopropylbenzyl Alcohol (I). To a refluxing suspension of 13 g (0.34 mol) of lithium aluminium hydride in 300 ml of tetrahydrofuran was added 100 g (0.675 mol) of isobutyrophenone (Aldrich) in 200 ml of tetrahydrofuran. After 2 hr of additional refluxing, the reaction mixture was cooled to room temperature and treated with 15 ml of water, 15 ml of 15% aqueous sodium hydroxide, and 39 ml of water. The suspension was filtered, the fil-
ter residue was washed thoroughly with 200 ml of tetrahydrofuran, and the solvent of the combined filtrates was evaporated under reduced pressure to yield 84.9 g of crude 1, which was purified by fractionation over a Vigreux column and gave 73.9 g (0.526 mol, 78% of theory) of 1, b p 77-78" (0.3 m m ) . N-(cu,cu-Dimethylphenethy1)acetamide(4a). T o a solution of 6.0 g (0.04 mol) of compound 1 in 6.0 ml of glacial acetic acid and 3.3 ml of 95% sulfuric acid was added dropwise 1.9 g (0.046 mol) of acetonitrile a t 70", and the mixture was kept for a n additional 1 hr at this same temperature. Then it was added to enough crushed ice to keep the resulting solution a t room temperature, treated with 14 ml of 25% aqueous sodium hydroxide, returned just to acidic by addition of 15% aqueous sulfuric acid, and extracted in two steps with 150 ml of methylene chloride. The combined extracts were washed with 3 X 50 ml of water and dried over sodium sulfate and the solvent was evaporated under reduced pressure to yield 7.1 g of an oil, which crystallized in cyclohexane to give 2.7 g (0.014 mol, 35% of theory) of 4a: m p 8990" (lit.3 m p 91-92"); nmr (CDC13) 2 CH3 a t 1.29 (s), CH3 at 1.85 (s), CH2 at 3.02 (s), CsH5 a t 7.19 ppm (complex). N-(a,a-Dimethylphenethy1)benzamide(4b). By the method described for the preparation of 4a, but using 3.3 g (0.032 mol) of benzonitrile instead of the acetonitrile, the crude product was 6.4 g of crystalline material which was purified by recrystallization in 95% ethanol: m p 109-110"; nmr (CDC13) 2 CH3 at 1.43 (s), CH2 a t 3.14 (s), 10 aromatic a t 7.1-7.8 ppm (complex). Anal. Calcd for C17H19NO: C, 80.60; H , 7.56; N , 5.53. Found: C, 80.95; H, 7.33; N, 5.57. N-(a-1sopropylbenzyl)acetamide ( l a ) . To a solution of 6.0 g (0.04 mol) of 1 and 1.9 g (0.046 mol) of acetonitrile in 6.0 ml of glacial acetic acid was added 3.3 ml of 95% sulfuric acid at 70". After the previously described work-up, which yielded 6.6 g of crude product, a recrystallization in 2-propanol gave 2.6 g (0.0135 mol, 3470 of theory) of 2a: m p 114-116"; nmr (CDC13) CH3 a t 0.78 (d), CH3 at 0.95 (d), CH3 a t 1.9 (s), CH at 1.9 (heptet), CH at 4.7 (broad triplet), CeH5 a t 7.2 ppm (s). Anal. Calcd for C12H17NO: C, 75.40; H, 8.96; N, 7.32. Found: C, 75.60; H, 8.56; Ti,7.37. N-(a-1sopropylbenzyl)benzamide (2b). By the method described for the preparation of 2a, b u t using 3.3 g (0.032 mol) of benzonitrile instead of acetonitrile, the crude product was 6.2 g of crystalline material, which was purified by recrystallization in ethanol: m p 139-140" (lit.4 m p 141"); nmr (CDC13) CH3 a t 0.87 (d), CH3 a t 1.02 (d), C H a t 2.23 (heptet), CH a t 5.01 (broad triplet), 10 aromatic a t 7.4-7.9 ppm (complex).
Acknowledgment. We wish to acknowledge the helpful discussions with Dr. J . L. Marsh and with Mr. L. Dorfman, whose staff we thank for microanalyses and spectra. Registry No.-1, 611-69-8; 2a, 33617-85-5; 2b, 51310-27-1; 4a, 5531-33-9;4b, 51310-28-2; isobutyrophenone, 611-70-1.
References and Notes ( 1 ) To w h o m inquiries should be directed to Dyestuffs & Chemicals Division, CIBA-GEIGY Corp.. P.O. Box 11422, Greensboro, N . C. 27409. ( 2 ) For a recent review of this reaction see L. i . Krimen and D. J. Cota, Org. Reacf., 17, 213 (19691. (3) K . - H . Boltze and H . Muhlenbein, German Patent 1,144,713 (Oct 1 4 ,
1960).
( 4 ) H . Christol, A. Laurent. and M . Mouseron, Buil. SOC. Chim. Fr.. 2313 (1961). (5) The structure of the products was assigned unequivocally from their nmr spectra. since the chirality of the benzylic carbon of 2 gives di-
astereotopic character to its two methyl groups.
N-Acyllactam Rearrangements. The Fate of the Carboxyl Carbon a n d t h e Synthesis of 2-tert-Butyl-1-pyrroline Bradford P. Mundy,* Kenneth B. Lipkowitz, Marc Lee, and Brent R. Larsen
Department of Chemistry, Montana State Uniuersity, Bozeman, Montana 59715 Received February 5, 1974
The rearrangement of N-acyllactams to 2-substituted cyclic imines has recently been investigated in our labora-
1964 J. Org. Chem., Vol. 39, No. 13, 1974 tory,lg2 and has resulted in a simple entry into pyrroline and piperideine alkaloids, 1-3 (Scheme I). As a result of some exploratory research into the possible role of la as an intermediate in nicotine b i o ~ y n t h e s i swe , ~ had occasion to consider the fate of the carboxyl carbon of the N-acyl moiety. Scheme I Synthesis of 2-Substituted Pyrrolines a n d Piperideine
Notes the tert-butyl group into a pyrroline, this seemed a reasonable test molecule for our synthetic efforts. Dry distillation of N-pivaloyl-2-pyrrolidone(6a) in the presence of calcium oxide gave 6b. The ease of preparing N-acyllactams, coupled with the specificity of the rearrangement regardless of ring size and constitution of R, suggest this to be a general and useful methodology for 2-substituted cyclic imines. Experimental Section
2a 3a
3b, n s 2; R = n-Pr (y-coniciene?
4a
4b, n-1; R
Sa 6a
5b, n-1: R=6b. n s 1; R = t-BU
=
u
To examine this point, and utilizing our experience with a readily available model, 14C-carboxyl labeled benzoyl chloride was treated with 2-pyrrolidone to yield 4a (specific activity 246 dpm/mmol). Dry distillation of an equal weight portion of 4a and calcium oxide resulted in formation of 4b (specific activity 233 dpm/mmol). Within the experimental error of our counting, one could suggest that all of the carbonyl carbon of the N-acyl group was maintained in the product. A partial mechanism consistent with this observation is delineated in Scheme 11. Based on a consideration of this type, one might predict that the reaction would not be sensitive to the inherent electronic and steric effects of the R group. Two further entries suggest this to be true, 5 and 6 (Scheme I). Scheme I1 Proposed Mechanism for N-Acyllactam Rearrangement
N-Cyclopropanoyl-2-pyrrolidone(5a) was readily con-
Nmr spectra were recorded on a Varian A-60 instrument, using T M S as an internal standard and deuteriochloroform as solvent. Boiling and melting points are uncorrected. Carbon-14 analyses were performed by liquid scintillation counting, using the channels ratio technique to determine efficiencies. Synthesis of 2-Phenyl-1-pyrroline (4b). Benzoyl chloride (140 g), 14C label a t the carbonyl carbon, was converted to N-benzoyl2-pyrrolidone by the method previously reported.1 The product was repeatedly crystallized from hot ethanol (mp 91") to constant specific activity of 246 dpm/mmol. This product was mixed with an equal weight of calcium oxide and the resulting mixture was heated with a free flame to give crude 2-phenyl-l-pyrr0line.~A sample was purified by preparative glc [lo ft x 0.25 in. column, 10% Carbowax 20M on Chromosorb W, impregnated with 6% (w/w) sodium hydroxide]. A sample counted 233 dpm/mmol. Synthesis of N-Cyclopropanoyl-2-pyrrolidone (5a). A solution of cyclopropanoyl chloride (5 g, 0.048 mol) and pyridine (10 g) was allowed to stir while 8.25 g of 2-pyrrolidone was slowly added. After stirring a t room temperature for 8 hr, the reaction mixture was taken up in methylene chloride and washed with dilute hydrochloric acid until the washings were acidic. The methylene chloride solution was washed with saturated sodium bicarbonate, and the resulting neutralized solution was dried over anhydrous sodium sulfate and reduced in volume. The crude product was crystallized from chloroform-hexane to yield 6.9 g (94%) of 5a: m p 62-63"; nmr 6 2.61 (t, 2), 1.97 (m, Z), 3.75 (t, 21, 3.17 (m, l ) ,0.97 (m, 4). Anal. Calcd for CXHllN02: C, 62.7; H, 7.2; N , 9.1. Found: C, 62.3; H , 7.1; N, 9.3. Synthesis of 2-Cyclopropyl-1-pyrroline (5b). Calcium oxide (1.5 g) was intimately mixed with an equal weight of 5a and the mixture was subjected to dry distillation with a free flame. Redistillation of the crude product yielded 0.78 g (73%) of 5b: bp 4.5' (10 m m ) ; picrate derivative m p 150-151"; nmr 6 3.75 ( m , 2), 0.81 (d, 4), 1.5-2.6 (unresolved multiplet, 5). Anal (picrate) Calcd for C13HI4N407: N, 16.5. Found: N, 16.5. Preparation of 2-tert-Butyl-1-pyrroline(6b). A solution of 2pyrrolidone (17 g) and pyridine (7.9 g) in 100 ml of benzene was stirred a t room temperature while pivaloyl chloride (12 g) in 25 ml of benzene was slowly added. The reaction mixture was diluted with ca. 100 ml of methylene chloride and washed with dilute hydrochloric acid until the washings were acidic. The methylene chloride layer was separated and washed with saturated hicarbonate. This solution was dried over anhydrous sodium carbonate, filtered. and reduced in volume. Distillation (62-65", 0.2 mm) gave 13.7 g (81% yield) of the desired N-acyllactam, which was taken immediately to the next step. An equal-weight mixture (6.72 g) of the above product and calcium oxide was heated with an open flame to yield the crude pyrroline. Distillation gave 2.0 g (40%) of 6b: bp 110" (1.5 m m ) ; nmr 6 1.15 (s, 9), 1.85 (m, 21, 2.5 (t, 2 ~ 3 . 7 5(t, 2j. Anal. Calcd for CBH15N: C. 76.8; H, 12.0; N, 11.2. Found: C, 76.8; H , 12.1; N, 10.9.
Acknowledgment. Acknowledgment is made t o the donors of the Petroleum Research Fund, administered by reorganization of the cyclopropyl moiety. The propensity the American Chemical Society, for partial support of this of this strained ring to suffer skeletal reorganization under many different reaction conditions is well d ~ c u m e n t e d . ~ research. For our last test of the nonparticipation of R during the Registry No.-4a, 2399-66-8: 4b, 700-91-4; 5a, 51269-66-0; 5b, reaction, we wished to examine a molecule which would 51269-67-1; 5b picrate, 51269-68-2; 6a, 51269-69-3; 6b, 51269-70-6; be difficult to prepare by other methods. Since our previcyclopropanoyl chloride, 4023-34-1; 2-pyrrolidone, 616-45-5; pivalous synthetic efforts had shown us that the sequence oyl chloride. 3282-30-2. worked equally well in the five- and six-ring series, our References and Notes most important consideration focused on the nature of the (1) 5 . P. Mundy. 5. R. Larsen, L. F. McKenzie, and G. Braden. J. Org. R group. A literature search revealed a synthesis of 2-tertChem., 37, 1635 (1972). (2) E. P. Mundyand 5 . R. Larsen, Syn. Commun., 2, 197 (1972). butyl-1-piperideine starting from 2-tert-butyl-1-pyridine .5 (3) Three N . tabaccum plants were administered ['LC]-y-aminobutyric Since no corresponding pyrroline had ever been reported, acid. After a metabolic period of 24 hr was allowed, the plants were frozen in liquid nitrogen and homogenized in chloroform. Nonlabeled and an obvious method was not apparent to incorporate verted t o 2-cyclopropyl-1-pyrroline(5b) without secondary
J. Org. Chem., Vol. 39, No. 13, 1974 1965
Notes l a was added to the homogenate with the expectatlon that any naturally occurring l a would mix with this and be reisolated. Careful purification of l a after reisolation demonstrated no activity to be present. This work was presented at the Montana Academy of Sciences,
Dillon, Mont., 1973. (4) See L. N . Ferguson in "Highlights of Alicyclic Chemistry," Franklin Publishing Corp., Palisade, N . J , , 1973, for a well-documented review on cyclopropane chemistry. (5) M . F . Brundon and B. E. Reynolds, J . Chem. Soc., 2445 (1964).
1
Br
Facile Addition of Bromine to a Reissert Compound T. George,* D. V. Mehta, and D. A. Dabholkar
2 Br
Contribution No. 366 from CIBA Research Centre, Goregaon, Bombay 63, India Received December 9, 1973 3
The chemistry of Reissert compounds has been reviewed Br in detail.1 Dihydro Reissert compounds have been described by Shamma, et al.;2 their nmr data have been studied by Bramley and J ~ h n s o n This . ~ note describes the facile addition of bromine to Reissert compound 1 and the tentative stereochemistry of the vicinal dibromide 2 and its derivatives. 4 The bromination of isoquinoline is reported to give 4Br bromoisoquinoline under drastic condition^.^ A mechanism involving the addition of bromine followed by the elimination of hydrogen bromide has been p ~ s t u l a t e d . ~ The dibromide has eluded isolation and characterisation. l-Cyano-2-p-toluenesulfonyl-1,2-dihydroisoquinoline(1) was prepared according to the method described by 5, R = OC2H, Wefer, et al.6 The bromination of 1 took place readily to 6, R = OCH(CHJ2 give l-cyano-3,4-cis-dibromo-2-p-toluenesulfonyl-l,2,3,47. R = O H tetrahydroisoquinoline (2), which could be isolated and characterized. The analytical values and the spectral data Br Br were in agreement with structure 2. Refluxing 2 with ethanol and 2-propanol gave 5 and 6, respectively. Treatment of 2 with aqueous sodium bicarbonate gave 4-bromo-lcyano-3-hydroxy-1,2,3,4-tetrahydroisoquinoline(7). The above a-amino alcohol 7 would be expected to exist in equilibrium with the open-chain structure 8 in solution. 9 However, poor solubility of the compound in solvents such from 2. However, the pathway involving the formation of as CHC13 and CC14 precluded such a study. The ir speciminium species 3 and loss of proton to give 4 cannot be trum of 7 in Nujol showed bands a t 3350 and 1700 cm-l, ruled out under the above conditions. Confirmation of the showing evidence for the &amino aldehyde structure 8. In product as 4 was obtained by establishing its identity with the nmr spectrum of the compound in CD3SOCD3 the sigthe product 'obtained by Reissert reaction on 4-bromoisonal corresponding to the aldehydic proton was not disquinoline. Refluxing 2 with 3 mol of morpholine in dioxcernible, showing the absence of any significant amount of ane caused elimination of p-toluenesulfonic acid along 8 in solution. with hydrogen bromide to give 4-bromoisoquinaldonitrile That the addition of bromine to 1 may take place in the (9). The latter compound has also been obtained by cis fashion is shown by the magnitude of the coupling conWefer, et a1.,6by use of sodium hydride in xylene. stant of C3 and C d protons in 2 ( J = 2.5 Hz). The above coupling constant is consistent also with a configuration Experimental Section in which the bromine atoms are trans and diaxial. HowMelting points are uncorrected. The ir spectra were examined ever, it is highly unlikely that the bromine atoms would as Nujol mulls on a Perkin-Elmer Model 421 spectrophotometer. be axial. The magnitude of the coupling constants of C3 The uv spectra in 95% ethanol were recorded on a Beckman and Cq protons in 5 , 6, and 7 remains of the same order as Model DK-2A spectrophotometer and the nmr spectra were recorded on a Varian Associates A-60 spectrometer with TMS as inin 2, indicating a cis orientation of the hydrogen atoms. ternal standard. The facile formation of the above compounds can be visul-Cyano-2-p-toluenesulfonyl-1,2-dihydroisoquinoline (1). The alised as the addition of elements of ethanol, 2-propanol, above compound was prepared in 52% yield according to the proor water to the iminium species 3 or the enamine 4. cedure of Wefer, et al .6 mp 101": ir 1620 and 1170 cm-I; uv max Treating 2 with 1 mol of morpholine in dioxane gave 4227 nm (log t 4.26) and 285 (3.93); nmr (CDC13) 6 2.31 (s, 3 H, bromo- 1-cyano - 2 - p-toluenesulfonyl-1,2-dihydroisoquino- ArCHd, 6.11 (d, 1 H, J = 6 Hz, Cq H),6.18 (d, 1 H, J = 0.6 Hz, line (4), in which the C1 proton occurs as a doublet (J = CI H), 6.80 (d, 1H, J = 6 Hz, Cs H), 6.95-7.90 (m, 8 H, aromatic). 3 ,I-cis-Dibromo- l-cyano-2-p-toluenesulfonyl-1,2,3,4-tetrahy0.5 Hz). This is in agreement with the observations of droisoquinoline (2). To a solution of 12.41 g (0.04 mol) of 1 in 150 Chhabra, e t ~ l . on , ~the long-range spin coupling in isoml of chloroform was added gradually 7.04 g (0.044 mol) of broquinoline Reissert compounds. In the presence of morphomine in 40 ml of chloroform at room temperature and the solution line, a facile trans-diaxial elimination of the elements of was stirred for 4 hr. Evaporation of the solvent gave a residue, hydrogen bromide can be postulated for the formation of 4 which crystallized on trituration with ether. It was recrystallized
